ALBERT

Description: Introduction: Approximately 40% of patients with Waldenström macroglobulinemia (WM) have an activating somatic mutation in CXCR4, including both nonsense and frameshift variants. There are limited data on the impact of CXCR4 mutations on the outcomes to therapy in WM patients. Mounting evidence suggests that CXCR4 mutations adversely affect depth of response and progression-free survival (PFS) to ibrutinib in patients with WM. Methods: We performed a pooled analysis evaluating the impact of CXCR4 mutations and CXCR4 mutation subtypes on response, PFS and survival after frontline treatment initiation (SAFTI) on 76 WM patients who received proteasome inhibitor-based primary therapy. All patients were participants on three prospective clinical trials evaluating the combinations of bortezomib, dexamethasone and rituximab (BDR; ClinicalTrials.Gov ID NCT00250926), carfilzomib, dexamethasone and rituximab (CaRD; ClinicalTrials.Gov ID NCT01470196), and ixazomib, dexamethasone and rituximab (IDR; ClinicalTrials.Gov ID NCT02400437) in previously untreated patients with WM, and were treated at the Bing Center for WM. All patients met criteria for a clinicopathological diagnosis of WM and for treatment initiation, according to the guidelines established by the 2nd International Workshop for WM (IWWM). CXCR4 mutations were divided in nonsense and frameshift mutations, as previously described. Response to therapy was assessed using response criteria by the 3rdIWWM. We fitted univariate and multivariate logistic regression and proportional-hazard Cox regression models for major response and PFS and SAFTI, respectively. P-values

Description: Abstract 2949 Poster Board II-925 Background: The efficacy of rituximab is dependent on a number of host immune interactions, including binding through excitatory (FcγRIIA, FcγRIIIA) as well as inhibitory (FcγRIIB) Fcγ receptors. In previous studies, we showed that polymorphisms in FcγRIIIA-158 predicted outcome to single agent rituximab therapy. Patients displaying L/H or L/R at FcγRIIIA-48 or at least one valine (V/V or V/F) at FcγRIIIA-158 demonstrated greater responses to rituximab versus those patients who expressed FcγRIIIA-48-L/L or FcγRIIIA-158 F/F, respectively (JCO 23:474). The predictive role of FcγRIIIA polymorphisms in patients receiving combination therapy with rituximab has not been addressed to date in WM. We therefore investigated the predictive role of FcγRIIIA-48, -158, as well as other important polymorphisms implicated in modulating IgG antibody binding and activation: FcγRIIA-27, -131, and FcγRIIB-187 in 65 patients with WM who received combination rituximab therapy. Patients and Methods: Sixty-four WM patients with a median age of 61, prior therapies of 0, IgM of 3,540 mg/dL, Hct of 32.3%, B2M of 2.7 g/L, who participated on a clinical study and whose outcomes have previously been reported were included in this analysis. Treatment included rituximab in combination with cyclophosphamide (n=43), thalidomide (n=14), or lenalidomide (n=7). Categorical responses for all patients were as follows: CR/VGPR 7 (11%); PR (n=30; 46.2%); MR (n=18; 27.7%); Non-Responders (n=9; 14.1%) for an overall response rate of 86%. Twenty seven patients have progressed with a median follow-up of 19.4 months. Polymorphic variants at FcγRIIA-27, -131, FcγRIIB-187, and FcγRIIIA-48, -158 were determined by Taq Man real time PCR analysis and sequencing, and impact on overall response, categorical response rates, and progression free survival determined. Results: The expression of H/H at FcγRIIA-131, or at least one valine (V/V or V/F) at FcγRIIIA-158 was associated with improved categorical response, particularly the attainment of CR/VGPR. For FcγRIIA-131, H/H was expressed in 2/9 (22.22%) WM patients who were non-responders; 13/38 (34.2%) patients attaining a major (≥ PR) response, and 4/7 (57.14%) patients who attained a CR/VGPR. For FcγRIIIA-158, the expression of at least one valine was observed in 3/9 (33.3%) WM patients who were non-responders; 20/38 (52.62%) patients attaining a major (≥ PR) response, and 5/7 (71.42%) patients who attained a CR/VGPR. Polymorphisms at FcγRIIA-27, and FcγRIIB-187 showed no association with response. The expression of L/H or L/R at FcγRIIIA-48 was observed in 3/7 (42.86%) patients with CR/VGPR, whereas 2/9 (22.22%) of patients who were non-responders expressed this polymorphism. Subset analysis showed that among patients who received cyclophosphamide based therapy, no differences in polymorphic variation for FcγRIIA-131, FcγRIIIA-48, and -158 were observed between non-responders, major responders and those achieving CR/VGPR. Conclusions: Taken together, the results of this study support a role for the use of FcγRIIA-131, FcγRIII-158, and possibly FcγRIIA-48 as determinants of better categorical responses in WM patients receiving combination therapy with an immunomodulatory agent. The combined use of cyclophosphamide with rituximab therapy appears to negate the inferior outcomes predicted by polymorphic variants in FcγRIIA-131, FcγRIIIA-48, and -158 which have previously been shown to be associated with lower response rates to single agent rituximab therapy. Disclosures: No relevant conflicts of interest to declare.

Description: Background: Hypogammaglobulinemia of the “uninvolved” immunoglobulins is commonly observed in Waldenstrom’s macroglobulinemia (WM), and has often been attributed to disease-related suppression. However, there is a paucity of information related to the pathogenesis of hypogammaglobulinemia in these disorders. Methods: We evaluated the incidence of IgA and IgG hypogammaglobulinemia in 207 patients with WM, and addressed the impact of therapy and response on IgA and IgG levels for 93 of these patients who required subsequent treatment. We also performed extensive sequence analysis of the promoter, all exonic, and flanking intronic regions from peripheral blood of 19 untreated WM patients who demonstrated IgA and/or IgG hypogammaglobulinemia for 8 genes often observed in common variable immunodeficiency disorders (CVID) and B cell deficiency i.e. AICDA, BTK, CD40, CD154, NEMO(IkBkG), TACI, SH2D1A, UNG. Results: At baseline, 120/207 (58.0%), 131/207 (63.3%), and 102/196 (49.3%) patients had abnormally low levels of serum IgG (

Description: Background: Waldenstrom Macroglobulinemia (WM) is characterized by widespread involvement of the bone marrow (BM), and lymphadenopathy in 20% of the patients, implying continuous trafficking of WM cells into and out of the BM and lymph nodes. The normal process of B-cell homing is regulated by cytokines, chemokines, and adhesion molecules. One of the most extensively studied chemokines in migration is stromal derived factor SDF-1 and its receptor CXCR4. Here we study the role of chemokine receptors, and the SDF-1/CXCR4 axis on migration and adhesion in WM. Methods: Flow cytometry for CXC and CC chemokine receptors (CXCR1, CXCR2, CXCR3, CXCR4, CXCR5, CCR2, CCR4, CCR5, CCR6 and CCR7), and adhesion molecules (VLA-4 and LFA-1) on WM cell lines (BCWM.1 and WM-WSU) and patient samples was performed. Migration was determined using the transwell migration assay (Costar, NY). Cells were placed in the upper chambers of the migration assay with 1% FCS medium in the presence of serial concentrations of SDF-1 in the lower chambers. After 4 hours of incubation, cells that migrated to the lower chambers were counted. Similarly, adhesion was determined using an adhesion assay (EMD Biosciences, San Diego, CA) with 96-well plated coated with fibronectin. Immunoblotting for proteins downstream of CXCR4 was performed. The CXCR4 inhibitor AMD3100 (10–100uM, Sigma, MO) and Gi protein inhibitor pertussis toxin PTX (10–200ng/ml, Sigma, MO) were used to inhibit CXCR4 signaling. Results: The following chemokine receptors were expressed on patient CD19+WM cells with over 30% expression: CXCR1 (mean 60%), CXCR2 (mean 47%), CXCR4 (mean 47%), CXCR5 (mean 69%), CCR4 (mean 54%) and CCR6 (mean 61%). Similar expression was observed on WM cell lines. We next determined the effect of SDF-1 on migration and signaling pathways in WM. SDF-1 (10–100nM) induced migration in a bell-shaped curve with 30nM inducing maximum migration (110% compared to control). SDF-1 30nM induced a rapid activation of signaling pathways downstream of CXCR4 including pERK1/2, pAKT, and pPKC at 1 min, with maximum activation at 5min. The CXCR4 inhibitor AMD3100 inhibited migration of BCWM.1 in the presence of 30nM SDF-1, with AMD3100 10uM inhibiting migration at 59% of control, and 20 to 50uM leading to a plateau in inhibition of migration at 54% of control. AMD3100 inhibited pERK and pPKC activation, downstream of CXCR4 in a dose-dependent fashion. Similar results were observed using PTX, with inhibition of migration of WM cells at 50% compared to control. To determine the role of SDF-1 on adhesion, we first demonstrated that WM cells from patients and cell lines expressed high levels of surface VLA-4 expression (mean 95% surface expression). WM cells had an increase in adhesion to fibronectin (VLA-4 ligand) compared to BSA control. AMD3100 10uM inhibited adhesion to fibronectin (63 % of control), indicating that the SDF-1/CXCR4 axis regulates adhesion. Conclusion: CXCR4 is highly expressed on WM cells and regulates migration and adhesion, indicating a potential role in regulating WM trafficking into the BM and lymph nodes. These studies provide the preclinical framework to study CXCR4 inhibitors in the regulation of homing and adhesion in WM.

Description: Key Points Transcription profiles associated with mutated MYD88, CXCR4, ARID1A, abnormal cytogenetics including 6q−, and familial WM are described. Mutated CXCR4 profiles show impaired expression of the tumor suppressor response induced by MYD88L265P and also G-protein/MAPK inhibitors.

Description: Background Waldenstrom's macroglobulinemia (WM) is an indolent non-Hodgkin's lymphoma characterized by the accumulation of IgM secreting lymphoplasmacytic cells in the bone marrow. CXCR4 is a chemokine receptor that promotes the survival, migration, and adhesion to the bone marrow stroma of WM lymphoplasmacytic cells (LPC) through interactions with its ligand CXCL12. Through whole genome sequencing, we identified somatic mutations in CXCR4 that affected 1/3 of WM patients. These mutations were identical or functionally similar to those associated with Warts, Hypogammaglobulinemia, Infection, and Myelokathexis (WHIM) syndrome (Hunter et al, ASCO 2012), a rare autosomal dominant genetic disorder that is caused by frame shift or nonsense mutations in the carboxyl-terminal cytoplasmic tail of CXCR4. In WHIM syndrome, loss of the c-terminal tail of CXCR4 impairs receptor internalization, thereby prolonging G-protein and β-arrestin signaling (Lugane et al, Blood 2008). Ibrutinib induces WM cell death, and is highly active in WM (Treon et al, ICML-12, 2013). Since the target of ibrutinib (BTK) is a known downstream target of CXCR4, we sought to clarify if ibrutinib activity in WM LPCs was modulated by WHIM-like mutations in CXCR4. Methods We first sought to confirm the frequency of WHIM-like mutations in 87 untreated WM patients by Sanger sequencing. The most common CXCR4 somatic mutation identified (S338X) in these studies was then cloned by PCR from CD19+ LPCs from a WM patient with this somatic mutation. Wild type (WT) and S338X CXCR4 cDNAs were subcloned into plenti-IRES-GFP vector, and transduced using an optimized lentiviral based strategy into BCWM.1 WM cells. Five days after transduction, GFP positive cells were sorted and used for functional studies. Surface expression of CXCR4 was determined by flow cytometeric analysis using a PE-conjugated anti-CXCR4 monoclonal antibody. The expression of phosphorylated BTK, AKT, and ERK1/2 was determined by western blot analysis. Cell proliferation was measured with alamar blue. Results Sanger sequencing identified nonsense or frame shift mutations (WHIM-like) in the c-terminal tail of CXCR4 in 28 of 87 (32%) patients, the most common of which was a non-sense mutation (S338X) that was present in 12 patients. BCWM.1 cells were then transduced with control vector, CXCR4 wild type or CXCR4 S338X mutant expressing vectors. Expression was confirmed by cDNA Sanger sequencing. Stably transduced cells exposed to ibrutinib (0.5uM or 1uM) showed significantly reduced cell proliferation (p

Description: CD40 ligand (CD40L) is a potent inducer of normal and malignant B-cell proliferation through interaction with CD40. We and others have observed excess mast cells (MC) in bone marrow (BM) biopsies of WM patients, which are commonly found admixed with tumor aggregates. (Tournilhac et al, JCO 2004, 22:571S). We therefore sought to clarify the role of MC in WM. Co-culture of 0.5% paraformaldehyde fixed, or sublethally irradiated HMC-1, LAD, and KU mast or basophilic cell lines and sorted BM lymphoplasmacytic cells (LPC) from 10 WM patients resulted in MC dose-dependent tumor colony formation and/or proliferation as assessed by 3H-thymidine uptake studies. As demonstrated by immunohistochemical, multicolor flow cytometric (CD117+FceRI+) and/or RT-PCR analysis, CD40L was expressed on BM MC from 29 of 31 (94%), 11 of 13 (85%), and 7 of 9 (78%) of WM patients, respectively. In contrast, cell surface CD40L expression was not detected by immunohistochemistry (p=0.00005) and flow cytometry (p=0.003) in 5 normal donors, and only faint expression for 1 of 5 normal donors by RT-PCR (p=0.09). Moreover, by multicolor flow cytometry, CD40 was expressed on BM tumor cells from 14/17 (83%) patients. CD40 functionality was confirmed either by the G28.5 CD40 agonistic antibody which induced dose dependent proliferation or by the rh-CD40L which partly prevented serum starvation-induced-apoptosis of WM LPC from 4/4 and 3/4 patients respectively. Importantly, expansion of tumor cells from 3 of 4 patients in mixed cultures with paraformaldehyde fixed MC was blocked in a dose dependent manner by use of a CD40L blocking protein (CD40:Fc). These studies demonstrate that CD40L is constitutively expressed on the cell surface of BM MC in WM and support the growth of WM tumor cells, and therefore provide the framework for therapeutic targeting of MC and MC-WM cell interactions in WM.

Description: Abstract 2870 Background: We and others have previously reported a paradoxical flare in serum IgM levels following rituximab and/or IVIG administration which can occur within hours of its administration, and can affect 40–70% of WM patients. The paradoxical flare can often lead to symptomatic hyperviscosity, and/or aggravation of other IgM mediated morbidities. Direct stimulation of WM cells by rituximab or IVIG does not lead to further IgM release, and IL-6 levels rise in those patients experiencing a rituximab induced IgM flare (Yang et al, ASH 2009). We therefore sought to clarify the molecular mechanisms permitting the rituximab or IVIG mediated IgM flare in WM. Patients and Method: We performed co-culture studies of WM cells utilizing a transwell system with granulocytes or monocytes in the presence or absence of rituximab or IVIG. Granulocyte and monocyte cell lysates and supernatants were collected after rituximab stimulation for further analysis. IL-6 levels were assessed by multiplex assays and real-time RT-PCR. siRNAs targeting FcgRIA, FcgRIB, FcgRIIA, FcgRIIB, and FcgRIIIA were used to knockdown Fcγ receptors on primary monocytes. Fcγ receptor expression following siRNA transfection was confirmed by real-time RT-PCR and flow cytometry. Proteins from Rituximab stimulated monocytes were analyzed with phospho-antibody arrays and confirmed with western blotting. Results: A significant increase in IgM release was observed following co-culture of primary WM cells with monocytes and co-incubation with either rituximab or IVIG; was associated with IL-6 release, and could be blocked by anti-IL6 antibodies. The induction of IL-6 following rituximab stimulation of monocytes was significantly reduced following siRNA knockdown of FcgRIIA. Phospho-antibody array and ingenuity pathway analyses revealed that both PI3K/AKT and MAPK pathways were involved in signaling initiated by rituximab stimulation of monocytes through FcgRIIA. Increased phosphorylation of PI3K, AKT, p38MAPK and ERK1/2 was confirmed by western blotting with phospho-specific antibodies. In addition, inhibitors specific for PI3K, AKT, p38MAPK and ERK1/2 also reduced IL-6 production. Conclusion: Taken together, the above data suggest that the rituximab and/or IVIG related IgM flare observed in WM patients is triggered by IL-6 in response to stimulation of bystander monocytes through FcgRIIA binding and activation of the PI3K/AKT and MAPK pathways. Inhibitors aimed at these pathways block IL-6 release from rituximab stimulated monocytes. The results provided a framework for investigation of pre-emptive therapies for blocking the IgM flare in WM patients being considered for rituximab and/or IVIG therapy. Disclosures: No relevant conflicts of interest to declare.

Description: The TNF ligand-receptor superfamily and their adaptor proteins regulate important B-cell signaling pathways, including CD40L-CD40 and APRIL/BLYS-TACI through adaptor protein TRAF2. These pathways promote B-cell differentiation and immunoglobulin heavy chain class switching. Defects in immunoglobulin heavy chain class switching and presence of constitutive IgA and IgG hypogammaglobulinemia in patients with WM have previously been reported by us (Hunter et al, ASH2006). In WM patients we identified several novel splice variants of TNF-family members CD40 and BLYS, and their adaptor protein TRAF2. Cloning, sequencing and alignment analysis document that aberrant splice transcripts of CD40, CD40-Va, Vb and Vc, and BLYS, BLYS-Va and Vb, result from partial exon skipping (CD40Va and Vc) and entire or partial intron retention (CD40-Vb, BLYS-Va and Vb), while the TRAF2 variant is a result of exon skipping only. Using RT-PCR DNA fragment analysis, malignant and normal B-cells from the bone marrow of 25 WM patients and 6 healthy donors (HDs) were screened for the expression of CD40, BLYS and TRAF2 splice transcripts. This analysis identified overexpression of CD40-Vb (15/25WM vs. 0/6HD; P=0.01), CD40-Vc (21/25WM vs. 0/6HD; P=0.0003), BLYS-Vb (18/25WM vs. 0/6HD; P=0.002) and TRAF2-V (21/25WM vs. 1/6HD; P=0.004) in WM patients. The expression of other splice variants CD40-Va (18/25WM vs.2/4HD; P=0.09) and BLYS-Va (23/25WM vs. 4/6HD P=0.2) in the same group of WM patients were not significant. We hypothesized that these aberrations are consequences of genetic variations (GVs) distributed in the vicinity of splicing elements of these genes, as well as, alterations may have occurred in the repertoire of splicing factors (SFs) with respect of their expression levels. To address these issues, we started sequencing CD40, BLYS, and TRAF2 gene segments that are subjected to aberrant splicing. Sequencing analysis of the CD40 gene from WM B cells revealed 6 recurrent genetic variations (GVs-defined as mutations occurring more than one patients) that include 2 missense (on exon 3) and 3 silent substitutions (on exons 3-4-5), and 1 frame-shift deletion on exon 5. These substitutions lead to amino acid changes on CD40 gene, while the frame-shift deletion may cause truncation of wild-type CD40 protein. We also identified recurrent GVs on introns 4 and 5. All these GVs detected on exons and introns are distributed in the vicinity of key splicing elements that create and/or activate a new splice site in precisely the position required for the splicing events to create CD40 variants. Also, using TaqMan low density array (TaqMan-LDA) we evaluated levels of major SFs and other TNF family members involved in CD40 and BLYS signaling. These analyses showed that patients expressing aberrant splice variants of CD40 and BLYS overexpress not only other members of TNF family but also major SFs: SF2/ASF (a proto-oncogene), U2, hNRPA1 and SRP55. TaqMan-LDA analysis suggests that these SFs may play a significant role in CD40, BLYS and TRAF2 splicing in WM patients since these transcripts were upregulated (1.2-2.2 fold higher) only in those patients which expressed CD40, BLYS and TRAF2 variants. In conclusion, presence of GVs in the vicinity of splicing elements and upregulation of SFs collectively may promote aberrant CD40, BLYS and TRAF2 splicing and thus modulate TNF family pathways supportive of B-cell differentiation and immunoglobulin heavy chain class switching in patients with WM.

Description: Waldenstrom’s macroglobulinemia is a B-cell disorder characterized by the infiltration of lymphoplasmacytic cells (LPC) in the bone marrow (BM), along with an IgM monoclonal gammopathy. Characteristic of WM is an increased number of mast cells (MC) which are found in association with LPC, and stimulate LPC growth through several TNF-family members including CD40L, APRIL and BLYS (Ann Oncol17:1275; Blood104:917A). As such, we have targeted MC in WM. One important growth and survival factor for MC is stem cell factor (SCF), which signals through CD117. Imatinib mesylate blocks SCF signaling through CD117, and induces apoptosis of WM BM MC and LPC, both of which highly express CD117 (Blood2004; 104:314b). As such, we performed this Phase II study of imatinib mesylate in patients with relapsed and refractory WM. Intended therapy consisted of imatinib mesylate which was initiated at 400 mg po qD over the first month, and subsequently dose escalated to 600 mg po qD for up to 2 years. Dose de-escalation to 300 mg po qD was permitted for toxicity. Twenty-eight patients were enrolled, 27 of whom are eligible for evaluation at final analysis. Median age was 61 (range 33–80 years), and median prior therapies was 2 (range 1–5). Twenty-four and four patients had relapsed and refractory disease, respectively. With a median follow-up of 6.3 months for all evaluable patients, median serum IgM levels declined from 3,110 to 2,530 mg/dL at best response (p=0.002). On an evaluable basis, 7/27 (26%) and 2/27 (7%) of patients attained a ≥25% and ≥50% decrease in serum IgM, respectively. Responses were prompt, and occurred at a median of 2.1 months. The median time to progression for responding patients was 8.4 (range 2–15 months). Hematological toxicities were the most common feature and ≥grade 2 toxicities included anemia (n=12), leucopenia (n=5), edema (n=3), thrombocytopenia (n=2) and neutropenia (n=1). Tryptase levels, which measure mast cell burden, declined in all patients for whom pre- and post-serum was available from 6.7 (range 2.2–10 ng/ml) to 3 (range 1–5.5 ng/ml) (p